> Merkliste Sie können Bookmarks mittels Listen verwalten, loggen Sie sich dafür bitte in Ihr SLUB Benutzerkonto ein.
Medientyp: E-Artikel Titel: Proposal of New Activation Function in Deep Image Prior Beteiligte: Segawa, Ryo; Hayashi, Hitoshi; Fujii, Shohei Erschienen: Wiley, 2020 Erschienen in: IEEJ Transactions on Electrical and Electronic Engineering Sprache: Englisch DOI: 10.1002/tee.23191 ISSN: 1931-4981; 1931-4973 Schlagwörter: Electrical and Electronic Engineering Entstehung: Anmerkungen: Beschreibung: <jats:p>We propose a new activation function and verify its performance in a deep image prior network. The new activation function is RSwish, which is Swish with a random slope for <jats:italic>x</jats:italic> < 0 assuming that the input value of the activation function is <jats:italic>x</jats:italic>. In addition, we manipulate the probability of the random number and observe the effect. As a result, we found that RSwish can perform better than Swish by manipulating probabilities according to the degree of color change in the image. © 2020 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.</jats:p>